AMD Instinct MI300 vs Intel Iris Xe Graphics 80EU Mobile Comparison

Where Each One Wins

The AMD Instinct MI300 and Intel Iris Xe Graphics 80EU Mobile occupy opposite extremes of the GPU spectrum, and the recorded data confirms they are not competitors in any practical sense. The MI300 is a compute accelerator built for dense, high-throughput workloads, while the Iris Xe is an integrated graphics solution designed for light rendering and display output. The benchmark results, where available, show zero recorded wins for either part because the head-to-head benchmark database contains no entries that directly compare these two products. That absence is itself informative: the two chips do not share a measurable performance envelope.

For compute-heavy tasks, the MI300 wins by default. Its FP32 throughput of 47.87 TFLOPS, texture rate of 1,496.0 GTexel/s, and 128 GB of HBM3 memory with 5.32 TB/s of bandwidth place it in a category that the Iris Xe cannot approach. The Iris Xe delivers 1.856 TFLOPS of FP32 performance, a 58.00 GTexel/s texture rate, and system-shared memory with bandwidth described only as system dependent. The MI300 also holds the edge in memory capacity and bandwidth by orders of magnitude, with 128 GB of dedicated HBM3 versus a shared pool that is not quantified in the database.

For mobile, low-power, and display-oriented tasks, the Iris Xe wins. Its 15 W TDP, integrated form factor, and support for DirectX 12 (12_1), OpenGL 4.6, and Vulkan 1.4 make it functional in portable devices where the MI300, with its 600 W TDP, 267 mm length, and no display outputs, cannot operate at all. The MI300 lists no APIs, no display outputs, and a pixel rate of 0 MPixel/s. The Iris Xe produces 29.00 GPixel/s and is marked as production active, while the MI300 has no production status recorded.

The data splits cleanly: the MI300 wins every raw compute metric, and the Iris Xe wins every usability metric. Neither part has any recorded benchmark wins against the other because no direct comparison exists in the database. The percentile versus all GPUs is 50 for both, an identical ranking that reflects the absence of benchmark scores rather than parity in performance.

Architecture Differences

The two chips are built on fundamentally different process nodes, foundries, and architectures. The MI300 uses AMD's CDNA 3.0 architecture on a 5 nm process at TSMC, while the Iris Xe uses Intel's Generation 12.2 architecture on a 10 nm process at Intel. The MI300's die is 1017 mm² with 153,000 million transistors and a transistor density of 150.4 million per mm². The Iris Xe has no recorded die size, transistor count, or density, but its process node is twice as large in linear terms, which indicates a much smaller and denser integration is not possible with the available data.

The chip designations differ as well. The MI300 uses the Aqua Vanjaram chip and belongs to the Instinct (MIx) generation, with the Radeon Instinct as its predecessor. The Iris Xe uses the Raptor Lake chip and belongs to the HD Graphics-M (Raptor Lake) generation, with Arc Graphics-M listed as its successor. The MI300 has no successor recorded, and the Iris Xe has no predecessor recorded.

The compute resources diverge sharply. The MI300 has 14,080 shading units, 880 texture mapping units, and 0 ROPs. The Iris Xe has 640 shading units, 40 TMUs, and 20 ROPs. The MI300 has no ray tracing cores or tensor cores listed, and neither does the Iris Xe. The MI300's FP32 and FP16 throughput are both 47.87 TFLOPS, indicating a 1:1 ratio, while the Iris Xe delivers 1.856 TFLOPS FP32 and 3.712 TFLOPS FP16, a 2:1 ratio. This means the Iris Xe can process half-precision data at twice the rate of full-precision data, whereas the MI300 treats both equally.

Memory architecture is entirely different. The MI300 uses 128 GB of HBM3 on an 8192-bit bus with 5.32 TB/s of bandwidth. The Iris Xe uses system-shared memory with a system-shared type, bus width, and system-dependent bandwidth. The MI300's memory clock is 1300 MHz with 5.2 Gbps effective data rate; the Iris Xe's memory clock is not independent, since it shares system memory. The MI300 uses a PCIe 5.0 x16 bus interface, while the Iris Xe connects through a Ring Bus.

Power and physical design also separate the two. The MI300 has a 600 W TDP, requires 2x 8-pin power connectors, and a suggested power supply of 1000 W. The Iris Xe has a 15 W TDP, no power connectors, and no suggested power supply, because it draws power from the host platform. The MI300 measures 267 mm in length and 111 mm in height, while the Iris Xe has no recorded dimensions and is classified as an IGP, meaning it is integrated into the processor package. The MI300 has no display outputs; the Iris Xe's display outputs are portable-device dependent.

Head-to-Head Benchmarks

The head-to-head benchmark table in the database is empty, with zero entries and zero wins for either side. No direct comparison can be made from recorded measurements. The nearest rivals lists for both products are also empty, so there are no adjacent scores or delta percentage values to anchor either part against a common competitor. The absence of data means every comparative statement must rely on the specification sheet alone.

The largest wins for the MI300 come from raw compute throughput. Its FP32 figure of 47.87 TFLOPS is 25.8 times the Iris Xe's 1.856 TFLOPS. Its texture rate of 1,496.0 GTexel/s is 25.8 times the Iris Xe's 58.00 GTexel/s. The MI300 has 14,080 shading units versus 640, a 22-fold difference. Memory bandwidth is 5.32 TB/s versus system dependent, and memory capacity is 128 GB versus system shared, neither of which can be expressed as a ratio from the available data.

The largest wins for the Iris Xe come from practical functionality. Its pixel rate of 29.00 GPixel/s contrasts with the MI300's 0 MPixel/s. Its API support includes DirectX 12 (12_1), OpenGL 4.6, and Vulkan 1.4, while the MI300 lists N/A for all three. Its 20 ROPs enable rasterization, something the MI300 cannot do with 0 ROPs. The Iris Xe's 15 W TDP is 40 times lower than the MI300's 600 W, making it usable in platforms where the MI300 would exceed power delivery limits.

Clock speeds differ as well. The MI300 runs at a base of 1000 MHz and boosts to 1700 MHz. The Iris Xe runs at a base of 300 MHz and boosts to 1450 MHz. The MI300's memory clock is 1300 MHz at 5.2 Gbps effective, while the Iris Xe's memory clock is system shared. The MI300's higher base and boost clocks contribute to its compute lead, but the architectural differences in shading units and memory bandwidth are the dominant factors.

Both parts share a release date of January 3, 2023, and neither has a launch MSRP recorded in the database. The MI300's percentile versus all GPUs is 50, and the Iris Xe's percentile is also 50, with both average benchmark scores at 0. These identical values confirm that neither part has been scored in the database, not that they perform equally.

The Verdict

The data supports a straightforward choice based on workload. The AMD Instinct MI300 is the compute accelerator for anyone processing large data sets, running FP16 or FP32 workloads at scale, or requiring massive memory bandwidth. Its 128 GB of HBM3, 5.32 TB/s of bandwidth, 47.87 TFLOPS of FP32, and 1,496.0 GTexel/s texture rate make it a dedicated throughput engine. It is not a graphics card in the conventional sense: no display outputs, no API support, no pixel output, and a 600 W power draw that demands a 1000 W power supply.

The Intel Iris Xe Graphics 80EU Mobile is the integrated solution for portable systems that need basic graphics output, video decode, and light 3D acceleration. Its 15 W TDP, 29.00 GPixel/s pixel rate, DirectX 12 (12_1), OpenGL 4.6, and Vulkan 1.4 support, and production-active status make it a functional component in laptops and compact devices. Its 640 shading units and 40 TMUs are modest but adequate for its intended role.

Neither part should be chosen for the other's job. The MI300 cannot drive a display, rasterize pixels, or run graphics APIs, so it fails as a consumer GPU. The Iris Xe cannot approach the MI300's compute density, memory capacity, or bandwidth, so it fails as a data-center accelerator. The empty head-to-head benchmark table reinforces that these products are not alternatives to each other. The database records no comparison because none is meaningful.

The identical percentile scores of 50 and average benchmark scores of 0 indicate that neither product has been entered into the benchmark pool. The verdict must therefore come from the specification sheet, and that sheet is unambiguous: the MI300 is for compute, the Iris Xe is for integrated graphics. The release date being the same for both, January 3, 2023, is a coincidence of scheduling and does not imply any relationship between the products.

FAQ

Q: Which GPU has higher FP32 performance?

A: The AMD Instinct MI300 delivers 47.87 TFLOPS of FP32 performance, while the Intel Iris Xe Graphics 80EU Mobile delivers 1.856 TFLOPS.

Q: Can the AMD Instinct MI300 output video to a display?

A: No. The MI300 has no display outputs and a pixel rate of 0 MPixel/s, so it cannot drive a monitor. The Iris Xe, by contrast, has portable-device dependent display outputs and a pixel rate of 29.00 GPixel/s.

Q: What memory does each GPU use?

A: The MI300 uses 128 GB of HBM3 memory on an 8192-bit bus with 5.32 TB/s of bandwidth. The Iris Xe uses system-shared memory with a system-shared type, bus width, and system-dependent bandwidth.

Q: What is the power draw of each GPU?

A: The MI300 has a 600 W TDP and requires 2x 8-pin power connectors with a suggested power supply of 1000 W. The Iris Xe has a 15 W TDP and no power connectors or suggested power supply.

Q: Which GPU supports more graphics APIs?

A: The Iris Xe supports DirectX 12 (12_1), OpenGL 4.6, and Vulkan 1.4. The MI300 lists N/A for all graphics APIs, meaning it has no graphics API support.

Q: When were both GPUs released?

A: Both the AMD Instinct MI300 and the Intel Iris Xe Graphics 80EU Mobile were released on January 3, 2023.

Specification Differences

| Specification | AMD Instinct MI300 | Intel Iris Xe Graphics 80EU Mobile |

|---|---|---|

| Architecture | CDNA 3.0 | Generation 12.2 |

| Process node | 5 nm | 10 nm |

| Foundry | TSMC | Intel |

| Transistors | 153,000 million | Not recorded |

| Die size | 1017 mm² | Not recorded |

| Transistor density | 150.4M / mm² | Not recorded |

| Base clock | 1000 MHz | 300 MHz |

| Boost clock | 1700 MHz | 1450 MHz |

| Memory clock | 1300 MHz, 5.2 Gbps effective | System Shared |

| Memory size | 128 GB | System Shared |

| Memory type | HBM3 | System Shared |

| Memory bus width | 8192 bit | System Shared |

| Memory bandwidth | 5.32 TB/s | System Dependent |

| Shading units | 14080 | 640 |

| TMUs | 880 | 40 |

| ROPs | 0 | 20 |

| Pixel rate | 0 MPixel/s | 29.00 GPixel/s |

| Texture rate | 1,496.0 GTexel/s | 58.00 GTexel/s |

| FP32 | 47.87 TFLOPS | 1.856 TFLOPS |

| FP16 | 47.87 TFLOPS (1:1) | 3.712 TFLOPS (2:1) |

| TDP | 600 W | 15 W |

| Slot width | Not recorded | IGP |

| Power connectors | 2x 8-pin | None recorded |

| Suggested PSU | 1000 W | None recorded |

| Bus interface | PCIe 5.0 x16 | Ring Bus |

| Display outputs | No outputs | Portable Device Dependent |

| DirectX | N/A | 12 (12_1) |

| OpenGL | N/A | 4.6 |

| Vulkan | N/A | 1.4 |

| Dimensions | 267 mm x 111 mm | Not recorded |

| Production status | Not recorded | Active |

| Predecessor | Radeon Instinct | None recorded |

| Successor | None recorded | Arc Graphics-M |

| Chip | Aqua Vanjaram | Raptor Lake |

| Generation | Instinct (MIx) | HD Graphics-M (Raptor Lake) |